📦 Memory Graph — 个人知识图谱
v1.0.0以 Markdown+YAML 形式持久保存的个人知识图谱,跨 Agent 共享,随时记录与召回用户生活、项目等上下文。
0· 232·1 当前·1 累计
下载技能包
最后更新
2026/3/20
安全扫描
OpenClaw
安全
high confidenceNULL
评估建议
This skill appears coherent and implements a local, file-based memory graph as described, but review the following before installing:
- Inspect the scripts yourself (they are bundled). They will read/write ~/memory/ and can create symlinks from the skill folder into ~/memory/scripts/. Ensure you trust the skill source before creating those symlinks.
- Node.js 22+ is required. Run scripts manually first (e.g., run rebuild-indexes.js, query.js) to confirm behavior before allowing any agent to inv...详细分析 ▾
✓ 用途与能力
Name/description (local file-based memory graph) match the files and scripts provided. The skill reads/writes ~/memory/, builds a local SQLite index, runs local Node.js scripts, and optionally commits to a local git repo — all expected for the stated purpose.
ℹ 指令范围
Runtime instructions and scripts operate on ~/memory/ (graph/, log/, indexes/, backfill/) and the local SQLite index. They do not reference external credentials or remote endpoints in the runtime scripts. Notable behaviors: scripts read and write local markdown files, rebuild indexes, and search logs; commit.js runs git commands (adds/stages/commits). This is within scope but means the skill will modify files and commit changes to the user's ~/memory repository when invoked.
✓ 安装机制
There is no install spec; this is instruction + bundled scripts. The setup process uses symlinks to expose scripts under ~/memory/scripts and requires Node.js 22+. No downloads from untrusted URLs or archive extraction are used.
✓ 凭证需求
The skill declares no required environment variables or credentials (scripts accept an optional MEMORY_ROOT). The scope of env access (home directory memory path) is appropriate for a local file-based memory graph.
ℹ 持久化与权限
always:false (normal). The skill's scripts write to and modify files under ~/memory/, and commit.js executes git commands in that directory. Allowing autonomous agent invocation (platform default) means an agent could autonomously run these scripts and modify your files; that increases blast radius but is coherent with the skill's purpose.
⚠ scripts/commit.js:28
Shell command execution detected (child_process).
⚠ scripts/rebuild-indexes.js:70
Shell command execution detected (child_process).
⚠ scripts/suggest-backfill.js:49
Shell command execution detected (child_process).
安全有层次,运行前请审查代码。
运行时依赖
无特殊依赖
版本
latestv1.0.02026/3/20
NULL
● 可疑
安装命令
点击复制官方npx clawhub@latest install agent-memory-graph
镜像加速npx clawhub@latest install agent-memory-graph --registry https://cn.longxiaskill.com